Proposes HAPS metric using frozen histopathology encoder with linear head, validated on expert-annotated H&E-IHC pairs and shown to improve virtual staining models by filtering low-similarity training data in the MIST dataset.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
HAPS: Rethinking Image Similarity for Virtual Staining
Proposes HAPS metric using frozen histopathology encoder with linear head, validated on expert-annotated H&E-IHC pairs and shown to improve virtual staining models by filtering low-similarity training data in the MIST dataset.